Beruflich Dokumente
Kultur Dokumente
, y
) = 0
This chapter is concerned with special yet very important second order equations, namely
linear equations.
Recall that a rst order linear dierential equation is an equation which can be written
in the form
y
+ p(x)y = q(x)
where p and q are continuous functions on some interval I. A second order, linear
dierential equation has an analogous form.
DEFINITION 1. A second order linear dierential equation is an equation which can
be written in the form
y
+p(x)y
5y
2x y
2
x
y
+
2
x
2
y = 0
where p(x) = 2/x, q(x) = 2/x
2
, f(x) = 0 are continuous on any interval that does
not contain x = 0. For example, we could take I = (0, ).
63
(c) y
+xy
2
y
y
3
= e
xy
is a nonlinear equation; this equation cannot be written in the
form (1).
Remarks on Linear. Intuitively, a second order dierential equation is linear if y
appears in the equation with exponent 1 only, and if either or both of y and y
appear
in the equation, then they do so with exponent 1 only. Also, there are no so-called cross-
product terms, y y
, y y
, y
+p(x)y
(x) +p(x)y
(x) + q(x)y(x),
then, for any two twice dierentiable functions y
1
(x) and y
2
(x),
L[y
1
(x) + y
2
(x)] = [y
1
(x) + y
2
(x)]
+p(x)[y
1
(x) +y
2
(x)]
+q(x)[y
1
(x) +y
2
(x)]
= y
1
(x) +y
2
(x) +p(x)[y
1
(x) + y
2
(x)] +q(x)[y
1
(x) +y
2
(x)]
= y
1
(x) +p(x)y
1
(x) +q(x)y
1
(x) +y
2
(x) +p(x)y
2
(x) + q(x) +y
2
(x)
= L[y
1
(x)] +L[y
2
(x)]
and, for any constant c,
L[cy(x)] = [cy(x)]
+ p(x)[cy(x)]
+q(x)[cy(x)]
= cy
(x) +p(x)[cy
(x)] +cq(x)y(x)
= c[y
(x) +p(x)y
(x) + q(x)y(x)]
= cL[y(x)].
Therefore, as introduced in Section 2.1, L is a linear dierential operator. This is the real
reason that equation (1) is said to be a linear dierential equation.
The rst thing we need to know is that an initial-value problem has a solution, and that
it is unique.
THEOREM 1. (Existence and Uniqueness Theorem) Given the second order linear
equation (1). Let a be any point on the interval I, and let and be any two real
numbers. Then the initial-value problem
y
+p(x) y
(a) =
has a unique solution.
64
As before, a proof of this theorem is beyond the scope of this course.
Remark: We can solve any rst order linear dierential equation; Section 2-1 gives a
method for nding the general solution of any rst order linear equation. In contrast, there
is no general method for solving second (or higher) order linear dierential equations. There
are, however, methods for solving certain special types of second order linear equations and
we shall study these in this chapter. Extensions of these methods to higher order linear
equations will be given later.
DEFINITION 2. The linear dierential equation (1) is homogeneous
1
if the function f
on the right side of the equation is 0 for all x I. In this case, equation (1) becomes
y
+p(x) y
+ q(x) y = 0. (2)
Equation (1) is nonhomogeneous if f is not the zero function on I, i.e., (1) is nonhomo-
geneous if f(x) = 0 for some x I.
As you will see in the work which follows, almost all of our attention will be focused on
homogeneous equations.
1
This use of the term homogeneous is completely dierent from its use to categorize the rst order
equation y
+ p(x) y
+ q(x) y = 0 (H)
where p and q are continuous functions on some interval I.
The trivial solution The rst thing to note is that the zero function, y(x) = 0 for all
x I, (also denoted by y 0) is a solution of (H) (y 0 implies y
0 and y
0).
The zero solution is called the trivial solution . Obviously our main interest is in nding
nontrivial solutions. Unless specied otherwise, the term solution will mean nontrivial
solution.
First we establish some essential facts about homogeneous equations.
THEOREM 1. If y = y(x) is a solution of (H) and if C is any real number, then
u(x) = Cy(x) is also a solution of (H).
Proof Let y = y(x) be a solution of (H). Then
y
(x) +p(x)y
(x) +q(x)y(x) = 0.
Let C be any real number, and set u(x) = Cy(x). Then
u(x) = Cy(x)
u
(x) = Cy
(x)
u
(x) = Cy
(x)
Substituting u into (H), we get
u
(x) +p(x) u
(x) + p(x)[Cy
(x)] +q(x)[Cy(x)]
= C[y
(x) +p(x) y
+ p(x)y
+ q(x)y.
Since y = y(x) is a solution of (H), L[y(x)] = 0. Since L is a linear operator,
L[Cy(x)] = C L[y(x)] = C(0) = 0.
Thus, u(x) = Cy(x) is a solution of (H).
In words, Theorem 1 says that any constant multiple of a solution of (H) is also a
solution of (H).
66
THEOREM 2. If y = y
1
(x) and y = y
2
(x) are any two solutions of (H), then
u(x) = y
1
(x) + y
2
(x) is also a solution of (H).
Proof Let y = y
1
(x) and y = y
2
(x) be any two solutions of (H). Then
y
1
(x) +p(x)y
1
(x) +q(x)y
1
(x) = 0 and y
2
(x) +p(x)y
2
(x) +q(x)y
2
(x) = 0.
Now set u(x) = y
1
(x) +y
2
(x). Then
u(x) = y
1
(x) +y
2
(x)
u
(x) = y
1
(x) +y
2
(x)
u
(x) = y
1
(x) +y
2
(x)
Substituting u into (H), we get
u
(x) +p(x) u
1
(x) +y
2
(x) +p(x)[y
1
(x) + y
2
(x)] +q(x)[y
1
(x) +y
2
(x)]
= [y
1
(x) + p(x)y
1
(x) + q(x)y
1
(x)] + [y
2
(x) +p(x)y
2
(x) +q(x)y
2
(x)]
= 0 + 0
= 0.
Alternate Proof Set L[y] = y
+p(x)y
3y
+ 2y = 0. (a)
We want to determine whether or not the two-parameter family
y = C
1
e
x
+C
2
e
2x
(b)
is the general solution of (a).
Let u = u(x) be any solution of (a) and let = u(0), = u
(0) = .
We have
y(x) = C
1
e
x
+C
2
e
2x
, y
(x) = C
1
e
x
+ 2C
2
e
2x
Setting x = 0, we obtain the pair of equations
y(0) = C
1
+ C
2
=
y
(0) = C
1
+ 2C
2
= .
This pair of equations has the unique solution C
1
= 2 , C
2
= . Now,
y = (2 )e
x
+ ( )e
2x
and u(x)
are each solutions of (a), and y(0) = u(0) = , y
(0) = u
2
x
y
+
2
x
2
y = 0. (a)
68
We want to determine whether or not the two-parameter family
y = C
1
x
2
+C
2
(5x
2
) (b)
is the general solution of (a).
By the Existence and Uniqueness Theorem there exists a unique solution u of (a) such
that u(1) = 1, u
(1) = 0
we obtain the pair of equations:
y(1) = C
1
+ 5C
2
= 1
y
(1) = 2C
1
+ 10C
2
= 0
There is no solution to this pair of equations. Therefore u is not a member of the two-
parameter family (b) and (b) is not the general solution of (a)
The problem here is that y
1
and y
2
are constant multiples of each other (y
2
=
5y
1
; or y
1
= y
2
/5). Notice that while (b) appears to be a two-parameter family, it is, in
fact, a one-parameter family:
y = C
1
x
2
+C
2
(5x
2
) = (C
1
+ 5C
2
)x
2
= Kx
2
.
You can verify that y
3
(x) = x is a solution of (a) which is dierent from y
1
(i.e.,
not a constant multiple of y
1
), and that
y = C
1
x +C
2
x
2
is the general solution of (a).
Lets consider the problem in general. Suppose that y = y
1
(x) and y = y
2
(x) are
solutions of equation (H). Under what conditions is (1) the general solution of (H)?
Let u = u(x) be any solution of (H) and choose any point a I. Suppose that
= u(a), = u
1
(a) +C
2
y
2
(a) =
If we multiply the rst equation by y
2
(a), the second equation by y
2
(a), and add, we
get
[y
1
(a)y
2
(a) y
2
(a)y
1
(a)]C
1
= y
2
(a) y
2
(a).
Similarly, if we multiply the rst equation by y
1
(a), the second equation by y
1
(a), and
add, we get
[y
1
(a)y
2
(a) y
2
(a)y
1
(a)]C
2
= y
1
(a) +y
1
(a).
69
We are guaranteed that this pair of equations has solutions C
1
, C
2
if and only if
y
1
(a)y
2
(a) y
2
(a)y
1
(a) = 0
in which case
C
1
=
y
2
(a) y
2
(a)
y
1
(a)y
2
(a) y
2
(a)y
1
(a)
and C
2
=
y
1
(a) +y
1
(a)
y
1
(a)y
2
(a) y
2
(a)y
1
(a)
.
Since a was chosen to be any point on I, we conclude that (2) is the general solution of
(H) if
y
1
(x)y
2
(x) y
2
(x)y
1
(x) = 0 for all x I.
DEFINITION 2. (Wronskian) Let y = y
1
(x) and y = y
2
(x) be solutions of (H).
The function W dened by
W[y
1
, y
2
](x) = y
1
(x)y
2
(x) y
2
(x)y
1
(x)
is called the Wronskian of y
1
, y
2
.
We use the notation W[y
1
, y
2
](x) to emphasize that the Wronskian is a function of x
that is determined by two solutions y
1
, y
2
of equation (H). When there is no danger of
confusion, well shorten the notation to W(x).
Remark There is a short-hand way to represent the Wronskian of two solutions of equation
(H) using a 2 2 determinant. Determinants will be dened and discussed in general in
Chapter 5. For now
W(x) =
y
1
(x) y
2
(x)
y
1
(x) y
2
(x)
= y
1
(x)y
2
(x) y
2
(x)y
1
(x).
Example 3. From Example 1, the functions y
1
(x) = e
x
and y
2
(x) = e
2x
are each
solutions of
y
3y
+ 2y = 0.
Their Wronskian is:
W(x) =
e
x
e
2x
e
x
2e
2x
= e
x
(2e
2x
) e
2x
(e
x
) = e
3x
= 0 for all x (, ).
From Example 2, the functions y
1
(x) = x
2
and y
2
(x) = 5x
2
are solutions of
y
2
x
y
+
2
x
2
y = 0.
Their Wronskian is:
W(x) =
x
2
5x
2
2x 10x
= x
2
(10x) 2x(5x
2
) = 10x
2
10x
2
0.
70
Also from Example 2, the functions y
1
(x) = x
2
and y
3
(x) = x are solutions of
y
2
x
y
+
2
x
2
y = 0.
Their Wronskian is:
W(x) =
x
2
x
2x 1
= x
2
(1) 2x(x) = x
2
= 0 for all x (0, ).
Here is the general result.
THEOREM 4. Let y = y
1
(x) and y = y
2
(x) be solutions of equation (H), and let
W(x) be their Wronskian. Exactly one of the following holds:
(i) W(x) = 0 for all x I; y
1
is a constant multiple of y
2
and vice versa.
(ii) W(x) = 0 for all x I and y = C
1
y
1
(x) + C
2
y
2
(x) is the general solution of (H)
Proof Let y = y
1
(x) and y = y
2
(x) be solutions of equation (H). Then
y
1
+ py
1
+qy
1
= 0 which implies y
1
= py
1
qy
1
y
2
+ py
2
+qy
2
= 0 which implies y
2
= py
2
qy
2
Set W = y
1
y
2
y
2
y
1
. Then
W
= y
1
y
2
+y
2
y
1
y
2
y
1
y
1
y
2
= y
1
y
2
y
2
y
1
= y
1
_
py
2
qy
2
y
2
_
py
1
qy
1
= p
_
y
1
y
2
y
2
y
= pW
which implies that
W
+ p(x)W = 0.
Therefore W is a solution of the rst order linear equation
y
+ p(x)y = 0.
Now, as we showed in Section 2.1,
W(x) = C e
p(x) dx
, for some constant C.
If C = 0, then W(x) = 0 for all x I; if C = 0, then W(x) = 0 for all x I.
We have already shown that if W(x) = 0 for all x I, then (1) is the general solution
of (H). We leave it as an exercise (Exercise 25) to show that if W 0 on I then y
1
is
a constant multiple of y
2
(and vice versa).
71
Example 4. Finishing Example 3, y = C
1
e
x
+ C
2
e
2x
is the general solution of
y
3y
+ 2y = 0;
y = C
1
x
2
+ C
2
x is the general solution of
y
2
x
y
+
2
x
2
y = 0.
DEFINITION 3. (Fundamental Set) A pair of solutions y = y
1
(x), y = y
2
(x) of
equation (H) forms a fundamental set of solutions if
W[y
1
, y
2
](x) = 0 for all x I.
Linear Dependence; Linear Independence
By Theorem 4, if y
1
and y
2
are solutions of equation (H) such that W[y
1
, y
2
] 0,
then y
1
is a constant multiple of y
2
. The question as to whether or not one function is
a multiple of another function and the consequences of this are of fundamental importance
in dierential equations and in linear algebra. We introduce the concept here; we will deal
with it in more generality later.
In this sub-section we are dealing with functions in general, not just solutions of the
dierential equation (H)
DEFINITION 4. (Linear Dependence; Linear Independence) Given two functions
f = f(x), g = g(x) dened on an interval I. The functions f and g are linearly dependent
on I if one of the functions is a constant multiple of the other. That is, f and g are
linearly dependent on I if there exists a number such that g(x) = f(x) for all x I,
or if there is a number such that f(x) = g(x) for all x I. The functions f and
g are linearly independent on I if they are not linearly dependent.
Remark The case where one of the functions is 0 is special: If either f or g is the
zero function, then f and g are linearly dependent. For example, suppose g 0, then
g = 0 f is a multiple of f.
The term Wronskian dened above for two solutions of equation (H) can be extended to
any two dierentiable functions f and g. Let f = f(x) and g = g(x) be dierentiable
functions on an interval I. The function W[f, g] dened by
W[f, g](x) = f(x)g
(x) g(x)f
(x)
is called the Wronskian of f, g.
There is a connection between linear dependence/independence and Wronskian.
72
THEOREM 5. Let f = f(x) and g = g(x) be dierentiable functions on an interval
I. If f and g are linearly dependent on I, then W(x) = 0 for all x I (W 0 on I).
Proof If f and g are linearly dependent on I, then there exists a number such
that g(x) = f(x) on I. Since g
(x) = f
(x), we have
W(x) = f(x)g
(x) g(x)f
(x) = f(x)[f
(x)] [f(x)]f
(x)
= f(x)f
(x) f(x)f
Let y = y
1
(x) and y = y
2
(x) be solutions of equation (H). Exactly one of
the following holds:
(i) W(x) = 0 for all x I; y
1
and y
2
are linear dependent.
(ii) W(x) = 0 for all x I; y
1
and y
2
are linearly independent and y = C
1
y
1
(x) +
C
2
y
2
(x) is the general solution of (H).
The statements y
1
(x), y
2
(x) forma fundamental set of solutions of (H) and y
1
(x), y
2
(x)
are linearly independent solutions of (H) are synonymous.
Exercises 3.2
Verify that the functions y
1
and y
2
are solutions of the given dierential equation. Do
they constitute a fundamental set of solutions of the equation?
1. y
6y = 0; y
1
(x) = e
3x
, y
2
(x) = e
2x
.
2. y
9y = 0; y
1
(x) = e
3x
, y
2
(x) = e
3x
.
3. y
+ 9y = 0; y
1
(x) = cos 3x, y
2
(x) = sin 3x.
4. y
4y
+ 4y = 0; y
1
(x) = e
2x
, y
2
(x) = xe
2x
.
5. x
2
y
x(x + 2)y
+ (x + 2)y = 0; y
1
(x) = x, y
2
(x) = xe
x
.
6. Given the dierential equation y
3y
4y = 0.
(a) Find two values of r such that y = e
rx
is a solution of the equation.
73
(b) Determine a fundamental set of solutions and give the general solution of the
equation.
(c) Find the solution of the equation satisfying the initial conditions y(0) =
1, y
(0) = 0.
7. Given the dierential equation y
_
2
x
_
y
_
4
x
2
_
y = 0.
(a) Find two values of r such that y = x
r
is a solution of the equation.
(b) Determine a fundamental set of solutions and give the general solution of the
equation.
(c) Find the solution of the equation satisfying the initial conditions y(1) =
2, y
(1) = 1.
(d) Find the solution of the equation satisfying the initial conditions y(2) = y
(2) =
0.
8. Given the dierential equation (x
2
+ 2x 1)y
2(x + 1)y
+ 2y = 0.
(a) Show that the equation has a linear polynomial and a quadratic polynomial as
solutions.
b Find two linearly independent solutions of the equation and give the general
solution.
Show that the given functions are linearly independent on the interval I and nd
a second-order linear homogeneous equation having the pair as a fundamental set of
solutions.
9. y
1
(x) = e
3x
, y
2
(x) = e
x
; I = (, ).
10. y
1
(x) = e
x
, y
2
(x) = xe
x
; I = (, ).
11. y
1
(x) = 1, y
2
(x) = x; I = (0, ).
12. y
1
(x) = cos 2x, y
2
(x) = sin 2x; I = (, ).
13. y
1
(x) = x, y
2
(x) = x
2
; I = (0, ).
14. y
1
(x) = x, y
2
(x) = x ln x; I = (0, ).
15. Let y = y
1
(x) be a solution of (H): y
+ p(x)y
t
a
p(u) du
y
2
1
(t)
dt.
Show that y
2
is a solution of (H) and that y
1
and y
2
are linearly independent.
Use Exercise 15 to nd a fundamental set of solutions of the given equation starting
from the given solution y
1
.
74
16. y
6y
+ 9y = 0; y
1
(x) = e
3x
.
17. y
2
x
y
+
2
x
2
y = 0; y
1
(x) = x.
18. y
1
x
y
+
1
x
2
y = 0; y
1
(x) = x.
19. y
1
x
y
4x
2
y = 0; y
1
(x) = e
x
2
.
20. y
2x 1
x
y
+
x 1
x
y = 0; y
1
(x) = e
x
.
21. Let y = y
1
(x) and y = y
2
(x) be solutions of equation (H) on an interval I. Let
a I and suppose that
y
1
(a) = , y
1
(a) = and y
2
(a) = , y
2
(a) = .
Under what conditions on , , , will the functions y
1
and y
2
be linearly
independent on I?
22. Suppose that the functions y
1
and y
2
are linearly independent solutions of (H).
Does it follow that c
1
y
1
and c
2
y
2
are also linearly independent solutions of (H)? If
not, why not.
23. Suppose that the functions y
1
and y
2
are linearly independent solutions of (H).
Prove that y
3
= y
1
+y
2
and y
4
= y
1
y
2
are also linearly independent solutions of
(H). Conversely, prove that if y
3
and y
4
are linearly independent solutions of (H),
then y
1
and y
2
are linearly independent solutions of (H).
24. Suppose that the functions y
1
and y
2
are linearly independent solutions of (H).
Under what conditions will the functions y
3
= y
1
+ y
2
and y
4
= y
1
+ y
2
be
linearly independent solutions of (H)?
25. Suppose that y = y
1
(x) and y = y
2
(x) are solutions of (H). Show that if y
1
(x) = 0
on I and W[y
1
, y
2
](x) 0 on I, then y
2
(x) = y
1
(x) on I.
75
3.3 Homogeneous Equations with Constant Coecients
We emphasized in Sections 3.1 and 3.2 that there are no general methods for solving second
(or higher) order linear dierential equations. However, there are some special cases for
which solution methods do exist. In this and the following sections we consider such a case,
linear equations with constant coecients. In this section we treat homogeneous equations;
nonhomogeneous equations will be treated in the next two sections.
A second order linear homogeneous dierential equation with constant coecients is an
equation which can be written in the form
y
+ ay
+by = 0 (1)
where a and b are real numbers.
You have seen that the function y = e
ax
is a solution of the rst-order linear equation
y
= r e
rx
and y
= r
2
e
rx
. Substitution into (1) gives
r
2
e
rx
+a (r e
rx
) +b (e
rx
) = e
rx
_
r
2
+ ar +b
_
= 0.
Since e
rx
= 0 for all x, we conclude that y = e
rx
is a solution of (1) if and only if
r
2
+ar +b = 0. (2)
Thus, if r is a root of the quadratic equation (2), then y = e
rx
is a solution of equation
(1); we can nd solutions of (1) by nding the roots of the quadratic equation (2).
DEFINITION 1. Given the dierential equation (1). The corresponding quadratic equa-
tion
r
2
+ar + b = 0
is called the characteristic equation of (1); the quadratic polynomial r
2
+ ar + b is
called the characteristic polynomial. The roots of the characteristic equation are called the
characteristic roots .
The nature of the solutions of the dierential equation (1) depends on the nature of the
roots of its characteristic equation (2). There are three cases to consider:
(1) Equation (2) has two, distinct real roots, r
1
= , r
2
= .
(2) Equation (2) has only one real root, r = .
76
(3) Equation (2) has complex conjugate roots, r
1
= + i , r
2
= i , = 0.
Case I: The characteristic equation has two, distinct real roots, r
1
= , r
2
= .
In this case,
y
1
(x) = e
x
and y
2
(x) = e
x
are solutions of (1). Since = , y
1
and y
2
are not constant multiples of
each other, the pair y
1
, y
2
forms a fundamental set of solutions of equation (1)
and
y = C
1
e
x
+ C
2
e
x
is the general solution.
Note: We can use the Wronskian to verify the independence of y
1
and y
2
:
W(x) =
e
x
e
x
e
x
e
x
= e
x
_
e
x
_
e
x
(e
x
) = () e
(+)x
= 0.
Example 1. Find the general solution of the dierential equation
y
+ 2y
8y = 0.
SOLUTION The characteristic equation is
r
2
+ 2r 8 = 0
(r + 4)(r 2) = 0
The characteristic roots are: r
1
= 4, r
2
= 2. The functions y
1
(x) = e
4x
, y
2
(x) = e
2x
form a fundamental set of solutions of the dierential equation and
y = C
1
e
4x
+ C
2
e
2x
is the general solution of the equation.
Example 2. Find a linearly independent pair of solutions of
y
+ 3y
= 0.
and give the general solution of the equation.
SOLUTION The characteristic equation is r
2
+3r = r(r +3) = 0, and the characteristic
roots are r
1
= 0, r
2
= 3. Therefore the functions y
1
(x) = e
0x
1 and y
2
(x) = e
3x
are linearly independent solutions of the dierential equation.
The general solution is
y = C
1
(1) + C
2
e
3x
= C
1
+ C
2
e
3x
.
77
Case II: The characteristic equation has only one real root, r = .
2
Then
y
1
(x) = e
x
and y
2
(x) = x e
x
are linearly independent solutions of equation (1) and
y = C
1
e
x
+C
2
x e
x
is the general solution.
Proof: We know that y
1
(x) = e
x
is one solution of the dierential equa-
tion; we need to nd another solution which is independent of y
1
. Since the
characteristic equation has only one real root, , the equation must be
r
2
+ar + b = (r )
2
= r
2
2r +
2
= 0
and the dierential equation (1) must have the form
y
2y
+
2
y = 0. ()
Now, z = C e
x
, C any constant, is also a solution of (), but z is not
independent of y
1
since it is simply a multiple of y
1
. We replace C by a
function u which is to be determined (if possible) so that y = ue
x
is a
solution of ().
3
Calculating the derivatives of y, we have
y = u e
x
y
= u e
x
+u
e
x
y
=
2
u e
x
+ 2u
e
x
+u
e
x
Substitution into () gives
2
u e
x
+ 2u
e
x
+u
e
x
2
_
u e
x
+u
e
x
+
2
u e
x
= 0.
This reduces to
u
e
x
= 0 which implies u
= 0 since e
x
= 0.
Now, u
e
x
xe
x
e
x
e
x
+ xe
x
= e
x
[e
x
+x e
x
] x e
x
= e
2x
= 0.
Finally, the general solution of () is
y = C
1
e
x
+ C
2
x e
x
.
Note: The solution y
2
(x) = xe
x
can also be obtained by using Problem 15
in Exercises 3.2.
Example 3. Find the general solution of the dierential equation
y
6y
+ 9y = 0.
SOLUTION The characteristic equation is
r
2
6r + 9 = 0
(r 3)
2
= 0
There is only one characteristic root: r
1
= r
2
= 3. The functions y
1
(x) = e
3x
, y
2
(x) = x e
3x
are linearly independent solutions of the dierential equation and
y = C
1
e
3x
+C
2
x e
3x
is the general solution.
Case III: The characteristic equation has complex conjugate roots:
r
1
= +i , r
2
= i , = 0
In this case
y
1
(x) = e
x
cos x and y
2
(x) = e
x
sin x
are linearly independent solutions of equation (1) and
y = C
1
e
x
cos x +C
2
e
x
sin x = e
x
[C
1
cos x + C
2
sin x]
is the general solution.
Proof: It is true that the functions z
1
(x) = e
(+i)x
and z
2
(x) = e
(i)x
are linearly independent solutions of (1), but these are complex-valued functions
79
and we are not equipped to handle such functions in this course. We want real-
valued solutions of (1). The characteristic equation in this case is
r
2
+ ar +b = (r [ + i ])(r [ i ]) = r
2
2r +
2
+
2
= 0
and the dierential equation (1) has the form
y
2y
+
_
2
+
2
_
y = 0. (*)
Well proceed in a manner similar to Case II. Set y = u e
x
where u is
to be determined (if possible) so that y is a solution of (*). Calculating the
derivatives of y, we have
y = u e
x
y
= u e
x
+u
e
x
y
=
2
u e
x
+ 2u
e
x
+u
e
x
Substitution into (*) gives
2
u e
x
+ 2u
e
x
+ u
e
x
2
_
u e
x
+u
e
x
+
_
2
+
2
_
u e
x
= 0.
This reduces to
u
e
x
+
2
u e
x
= 0 which implies u
+
2
u = 0 since e
x
= 0.
Now,
u
+
2
u = 0
is the equation of simple harmonic motion (for example, it models the oscillatory
motion of a weight suspended on a spring). The functions u
1
(x) = cos x and
u
2
(x) = sin x form a fundamental set of solutions. (Verify this.)
Since y = u e
x
, we conclude that
y
1
(x) = e
x
cos x and y
2
(x) = e
x
sin x
are solutions of (*). Its easy to see that y
1
and y
2
form a fundamental set of
solutions. This can also be checked by using the Wronskian
Finally, we conclude that the general solution of equation (1) is:
y = C
1
e
x
cos x + C
2
e
x
sin x = e
x
[C
1
cos x + C
2
sin x] .
Example 4. Find the general solution of the dierential equation
y
4y
+ 13y = 0.
80
SOLUTION The characteristic equation is: r
2
4r + 13. By the quadratic formula, the
roots are
r
1
, r
2
=
(4)
_
(4)
2
4(1)(13)
2
=
4
16 52
2
=
4
36
2
=
4 6 i
2
= 2 3 i.
The characteristic roots are the complex numbers: r
1
= 2 + 3 i, r
2
= 2 3 i. The
functions y
1
(x) = e
2x
cos 3x, y
2
(x) = e
2x
sin 3x are linearly independent solutions of the
dierential equation and
y = C
1
e
2x
cos 3x +C
2
e
2x
sin 3x = e
2x
[C
1
cos 3x +C
2
sin 3x]
is the general solution.
Example 5. Find two linearly independent solutions of y
+ 16y = 0.
SOLUTION The characteristic equation is r
2
+ 16 = 0 and the complex numbers r
1
=
0 + 4i = 4i, r
2
= 0 4i = 4i are the characteristic roots. The functions
y
1
(x) = e
0x
cos 4x = cos 4x, y
2
(x) = e
0x
sin 4x = sin 4x
are linearly independent solutions of the dierential equation.
In our next example we nd the solution of an initial-value problem.
Example 6. Find the solution of the initial-value problem:
y
+ 2y
15y = 0, y(0) = 2, y
(0) = 6.
SOLUTION The characteristic equation is
r
2
+ 2r 15 = 0
(r + 5)(r 3) = 0
The characteristic roots are: r
1
= 5, r
2
= 3. The functions y
1
(x) = e
5x
, y
2
(x) = e
3x
are linearly independent solutions of the dierential equation and
y = C
1
e
5x
+ C
2
e
3x
is the general solution.
Before applying the initial conditions we need to calculate y
:
y
(x) = 5C
1
e
5x
+ 3C
2
e
3x
Now, the conditions y(0) = 2, y
+y
6y = 0.
Example 8. Find a second order, linear, homogeneous dierential equation with constant
coecients that has y = C
1
e
4x
+ C
2
x e
4x
as its general solution.
SOLUTION Since e
4x
and xe
4x
are solutions, 4 must be a double root of the
characteristic equation. Therefore, the characteristic equation is
(r [4])
2
= (r + 4)
2
= 0 which expands to r
2
+ 8r + 16 = 0
and the dierential equation is
y
+ 8y
+ 16y = 0.
Example 9. Find a second order, linear, homogeneous dierential equation with constant
coecients that has y(x) = e
x
cos 2x as a solution.
SOLUTION Since e
x
cos 2x is a solution, the characteristic equation must have the
complex numbers 1 + 2i and 1 2i as roots. (Although we didnt state it explicitly,
e
x
sin 2x must also be a solution.) The characteristic equation must be
(r [1 + 2i])(r [1 2i]) = 0 which expands to r
2
2r + 5 = 0
and the dierential equation is
y
2y
+ 5y = 0.
82
Exercises 3.3
Find the general solution of the given dierential equation.
1. y
+ 2y
8y = 0.
2. y
13y
+ 42y = 0.
3. y
10y
+ 25y = 0.
4. y
+ 2y
+ 5y = 0.
5. y
+ 4y
+ 13y = 0.
6. y
= 0.
7. y
+ 2y
= 0.
8. 2y
+ 5y
3y = 0.
9. y
12y = 0.
10. y
+ 12y = 0.
11. y
2y
+ 2y = 0.
12. y
3y
+
9
4
y = 0.
13. y
30y = 0.
14. 2y
+ 3y
= 0.
15. 2y
+ 2y
+ y = 0.
16. y
+ 2y
+ 3y = 0.
17. y
8y
+ 16y = 0.
18. 5y
+
11
4
y
3
4
y = 0.
Find the solution of the initial-value problem.
19. y
5y
+ 6y = 0; y(0) = 1, y
(0) = 1.
20. y
+ 4y
+ 3y = 0; y(0) = 2, y
(0) = 1.
21. y
+ 2y
+y = 0; y(0) = 3, y
(0) = 1.
22. y
+
1
4
y = 0; y() = 1, y
() = 1.
23. y
2y
+ 2y = 0; y(0) = 1, y
(0) = 1.
83
24. y
+ 4y
+ 4y = 0; y(1) = 2, y
(1) = 1.
Find a dierential equation y
+ay
+ ay
2y = 0; y(0) =
, y
y = 0; y(0) = 2, y
(0) =
. Then nd such that y(x) 0 as x .
36. Given the dierential equation y
(2a 1)y
+ a(a 1)y = 0.
(a) Determine the values of a (if any) for which all solutions have limit 0 as
x .
(b) Determine the values of a (if any) for which all solutions are unbounded as
x .
Exercises 37 - 39 are concerned with the dierential equation (1): y
+ ay
+ by = 0
where a and b are constants.
37. Give a condition on a and b which will imply that:
(a) (1) has solutions of the form y
1
= e
x
, y
2
= e
x
, , distinct real numbers.
(b) (1) has solutions of the form y
1
= e
x
, y
2
= xe
x
, a real number.
(c) (1) has solutions of the form y
1
= e
x
cos x, y
2
= e
x
sin x, , real
numbers.
38. Prove that if a and b are both positive, then all solutions have limit 0 as x .
84
39. Prove:
(a) If a = 0 and b > 0, then all solutions of the equation are bounded.
(b) If a > 0 and b = 0, and y = y(x) is a solution, then
lim
x
y(x) = k for some constant k.
Determine k for the solution that satises the initial conditions y(0) =
, y
(0) = .
40. Show that the general solution of the dierential equation
y
2
y = 0, a positive constant,
can be written
y = C
1
cosh x +C
2
sinh x.
41. Suppose that the roots r
1
, r
2
of the characteristic equation (2) are real and distinct.
Then they can be written as r
1
= +, r
2
= where and are real. Show
that the general solution of equation (1) in this case can be expressed in the form
y = e
x
(C
1
cosh x + C
2
sinh x) .
Euler Equations A second order linear homogeneous equation of the form
x
2
d
2
y
dx
2
+x
dy
dx
+ y = 0 (E)
where and are constants, is called an Euler equation .
42. Prove that the Euler equation (E) can be transformed into the second order equation
with constant coecients
d
2
y
dz
2
+a
dy
dz
+by = 0
where a and b are constants, by means of the change of independent variable
z = ln x.
Find the general solution of the Euler equations.
43. x
2
y
xy
8y = 0.
44. x
2
y
2xy
+ 2y = 0.
45. x
2
y
3xy
+ 4y = 0.
46. x
2
y
xy
+ 5y = 0.
85
3.4 Second Order Linear Nonhomogeneous Equations
In this section we consider the general second order linear nonhomogeneous equation
y
+ p(x)y
+p(x)y
+q(x)y = 0. (H)
As we shall see, there is a close connection between equations (N) and (H). In this
context, equation (H) is called the reduced equation of equation (N).
General Results
THEOREM 1. If z = z
1
(x) and z = z
2
(x) are solutions of equation (N), then
y(x) = z
1
(x) z
2
(x)
is a solution of equation (H).
Proof: Since z
1
and z
2
are solutions of (N),
z
1
(x) +p(x)z
1
(x) +q(x)z
1
(x) = f(x) and z
2
(x) +p(x)z
2
(x) + q(x)z
2
(x) = f(x).
Let y(x) = z
1
(x) z
2
(x). Then
y
py
+qy = (z
1
z
2
) +p(z
1
z
2
) +q(z
1
z
2
)
=
_
z
1
+pz
1
+qz
1
_
_
z
2
+pz
2
+qz
2
_
= f(x) f(x) = 0.
Thus, y = z
1
z
2
is a solution of (H).
Alternate Proof Set L[y] = y
+ p(x)y
+p(x)y
+p(x)y
+p(x)y
+ q(x)y = g(x),
respectively, then z(x) = z
f
(x) +z
g
(x) is a particular solution of ().
The proof is left as an exercise.
This result can be extended to nonhomogeneous equations whose right-hand side is the
sum of an arbitrary number of functions.
87
COROLLARY If
z = z
1
(x) is a particular solution of
y
+ p(x)y
+q(x)y = f
1
(x),
z = z
2
(x) is a particular solution of
y
+ p(x)y
+q(x)y = f
2
(x),
.
.
.
z = z
n
(x) is a particular solution of
y
+ p(x)y
+q(x)y = f
n
(x),
then z(x) = z
1
(x) + z
2
(x) + + z
n
(x) is a particular solution of
y
+ p(x)y
+q(x)y = f
1
(x) + f
2
(x) + +f
n
(x).
The importance of Theorem 3 and its Corollary is that we need only consider non-
homogeneous equations in which the function on the right-hand side consists of one term
only.
Variation of Parameters
By our work above, to nd the general solution of (N) we need to nd:
(i) a linearly independent pair of solutions y
1
, y
2
of the reduced equation (H), and
(ii) a particular solution z of (N).
The method of variation of parameters uses a pair of linearly independent solutions of
the reduced equation to construct a particular solution of (N).
Let y
1
(x) and y
2
(x) be linearly independent solutions of the reduced equation
y
+p(x)y
+q(x)y = 0. (H)
Then
y = C
1
y
1
(x) +C
2
y
2
(x)
is the general solution of (H). We replace the arbitrary constants C
1
and C
2
by functions
u = u(x) and v = v(x), which are to be determined so that
z(x) = u(x)y
1
(x) +v(x)y
2
(x)
88
is a particular solution of the nonhomogeneous equation (N). The replacement of the pa-
rameters C
1
and C
2
by the variables u and v is the basis for the term variation of
parameters. Since there are two unknowns u and v to be determined, we shall impose
two conditions on these unknowns. One condition is that z should solve the dierential
equation (N). The second condition is at our disposal and we shall choose it in a manner
that will simplify our calculations.
Dierentiating z we get
z
= u y
1
+ y
1
u
+ v y
2
+y
2
v
.
For our second condition on u and v, we set
y
1
u
+y
2
v
= 0. (a)
This condition is chosen because it simplies the rst derivative z
becomes
z
= u y
1
+ v y
2
(b)
and
z
= u y
1
+ y
1
u
+v y
2
+y
2
v
.
Now substitute z, z
+pz
+ qz = (u y
1
+ y
1
u
+ v y
2
+ y
2
v
) + p(u y
1
+v y
2
) + q(u y
1
+ v y
2
)
= u(y
1
+py
1
+ qy
1
) + v(y
2
+ py
2
+qy
2
) + y
1
u
+y
2
v
.
Since y
1
and y
2
are solutions of (H),
y
1
+py
1
+ qy
1
= 0 and y
2
+ py
2
+qy
2
= 0
and so
z
+pz
+ qz = y
1
u
+y
2
v
.
The condition that z should satisfy (N) is
y
1
u
+y
2
v
= f(x). (c)
Equations (a) and (c) constitute a system of two equations in the two unknowns u and
v:
y
1
u
+ y
2
v
= 0
y
1
u
+ y
2
v
= f(x)
89
Obviously this system involves u
and v
and
v
and v
, we nd that
u
=
y
2
f
y
1
y
2
y
2
y
1
and v
=
y
1
f
y
1
y
2
y
2
y
1
We know that the denominators here are non-zero because the expression
y
1
(x)y
2
(x) y
2
(x)y
1
(x) = W(x)
is the Wronskian of y
1
and y
2
, and y
1
, y
2
are linearly independent solutions of the
reduced equation.
We can now get u and v by integrating:
u =
_
y
2
(x)f(x)
W(x)
dx and v =
_
y
1
(x)f(x)
W(x)
dx.
Finally
z(x) = y
1
(x)
_
y
2
(x)f(x)
W(x)
dx +y
2
(x)
_
y
1
(x)f(x)
W(x)
dx (2)
is a particular solution of the nonhomogeneous equation (N).
Remark This result illustrates why the emphasis is on linear homogeneous equations. To
nd the general solution of the nonhomogeneous equation (N) we need a fundamental set
of solutions of the reduced equation (H) and one particular solution of (N). But, as we
have just shown, if we have a fundamental set of solutions of (H), then we can use them to
construct a particular solution of (N). Thus, all we really need to solve (N) is a fundamental
set of solutions of its reduced equation (H).
Example 1. Find a particular solution of the nonhomogeneous equation
y
2
x
y
+
2
x
2
y = 2x
3
()
given that y
1
(x) = x and y
2
(x) = x
2
are linearly independent solutions of the correspond-
ing reduced equation. Also give the general solution of the nonhomogeneous equation.
SOLUTION The Wronskian of y
1
, y
2
is W(x) = y
1
y
2
y
2
y
1
= x(2x) x
2
(1) = x
2
.
By the method of variation of parameters, a particular solution of the nonhomogeneous
equation is
z(x) = u(x) x + v(x) x
2
where, from (2),
u(x) =
_
y
2
(x) f(x)
W(x)
dx =
_
x
2
(2x
3
)
x
2
dx =
_
2x
3
dx =
1
2
x
4
90
and
v(x) =
_
y
1
(x) f(x)
W(x)
dx =
_
x(2x
3
)
x
2
dx =
_
2x
2
dx =
2
3
x
3
(NOTE: Since we are seeking only one function u and one function v we have not included
arbitrary constants in the integration steps.)
Now
z(x) =
1
2
x
4
x +
2
3
x
3
x
2
=
1
6
x
5
.
is a particular solution of the nonhomogeneous equation () and
y = C
1
x + C
2
x
2
+
1
6
x
5
.
is the general solution.
Remark Rather than simply memorizing the formula (4) for the particular solution z of
(N), some people prefer to use the variation of parameters method to construct the solution
z. Well use this approach in the next example.
Example 2. Find the general solution of
y
5y
+ 6y = 4e
2x
. ()
SOLUTION The reduced equation y
5y
= 2ue
2x
+ e
2x
u
+ 3ve
3x
+e
3x
v
.
Imposing the condition
e
2x
u
+e
3x
v
= 0, (a)
the rst derivative simplies to
z
= 2ue
2x
+ 3ve
3x
(b)
and
z
= 4ue
2x
+ 2e
2x
u
+ 9ve
3x
+ 3e
3x
v
.
91
Substituting z, z
+ 9ve
3x
+ 3e
3x
v
5(2ue
2x
+ 3ve
3x
) + 6(u
2x
+ ve
3x
) = 2e
2x
u
+ 3e
3x
v
.
Setting this equal to 4e
2x
we have our second equation
2e
2x
u
+ 3e
3x
v
= 4e
2x
. (c)
Taking (a) and (c) together, we get the system
e
2x
u
+ e
3x
v
= 0
2e
2x
u
+ 3e
3x
v
= 4e
2x
Multiplying the rst equation by 3 and adding gives
e
2x
u
= 4e
2x
which implies u
= 4 and u = 4x;
multiplying the rst equation by 2 and adding gives
e
3x
v
= 4e
2x
which implies v
= 4e
x
and v = 4e
x
.
Therefore,
z = (4x)e
2x
+ (4e
x
)e
3x
= 4xe
2x
4e
2x
is a particular solution of (*).
Finally, the general solution of (*) is
y = C
1
e
2x
+C
2
e
3x
4x e
2x
4e
2x
which can be written equivalently as
y = C
1
e
2x
+ C
2
e
3x
4x e
2x
by combining 4e
2x
with C
1
e
2x
.
Exercises 3.4
Verify that the given functions y
1
and y
2
form a fundamental set of solutions of the
reduced equation of the given nonhomogeneous equation; then nd a particular solution of
the nonhomogeneous equation and give the general solution of the equation.
1. y
2
x
2
y = 3 x
2
; y
1
(x) = x
2
, y
2
(x) = x
1
.
2. y
1
x
y
+
1
x
2
y =
2
x
; y
1
(x) = x, y
2
(x) = x ln x.
92
3. x
2
y
2xy
+ 2y = x
2
ln x; y
1
(x) = x, y
2
(x) = x
2
.
4. y
1 + x
x
y
+
1
x
y = xe
2x
; y
1
(x) = 1 + x, y
2
(x) = e
x
.
5. (x 1)y
xy
+y = (x 1)
2
; y
1
(x) = x, y
2
(x) = e
x
.
6. x
2
y
xy
+y = 4x ln x.
Find the general solution of the given nonhomogeneous dierential equation.
7. y
2y = 2e
x
.
8. y
+y = tan x.
9. y
+ 4y = sec 2x.
10. y
2y
+y = xe
x
.
11. y
2y
+y = e
x
cos x.
12. y
4y
+ 4y =
1
3
x
1
e
2x
.
13. y
+ 4y
+ 4y =
e
2x
x
2
.
14. y
+ 2y
+y = e
x
ln x.
15. y
+ 9y = 9 sec
2
3x.
16. y
2y
+ 2y = e
x
sec x.
17. The function y
1
(x) = x is a solution of x
2
y
+xy
+ xy
y = 2x.
HINT: See Exercise 15, Section 3.2.
18. The function y
1
(x) = x is a solution of (x
2
+1)y
2xy
2xy
+ 2y = (x
2
+ 1)
2
.
19. The functions y
1
(x) = x
2
+x ln x, y
2
(x) = x +x
2
and y
3
(x) = x
2
are solutions of
a second order linear nonhomogeneous equation. What is the general solution of the
equation?
20. The functions y
1
(x) = x2x
3
y
2
(x) = xe
x
+x2x
3
and y
3
(x) = 2x
3
are solutions
of a second order linear nonhomogeneous equation. What is the general solution of
the equation?
93
3.5 Nonhomogeneous Equations with Constant Coecients;
Undetermined Coecients
Solving a linear nonhomogeneous equation depends, in part, on nding a particular solution
of the equation. We have seen one method for nding a particular solution, the method
of variation of parameters. In this section we present another method, the method of
undetermined coecients.
Remark: Limitations of the method. In contrast to variation of parameters, which
can be applied to any nonhomogeneous equation, the method of undetermined coecients
can be applied only to nonhomogeneous equations of the form
y
+ay
+ ay
+ by. (2)
If we calculate L[y] for an exponential function z = Ae
rx
, A a constant, we have
z = Ae
rx
, z
= Are
rx
, z
= Ar
2
e
rx
and
L[y] = y
+ ay
+ by = Ar
2
e
rx
+ a(Are
rx
) +b(Ae
rx
=
_
Ar
2
+aAr + bA
_
e
rx
= Ke
rx
where K = Ar
2
+ aAr +bA.
That is, the operator L transforms Ae
rx
into a constant multiple of e
rx
. We can use
this result to determine a particular solution of a nonhomogeneous equation of the form
y
+ ay
+ by = ce
rx
.
Here is a specic example.
Example 1. Find a particular solution of the nonhomogeneous equation
y
2y
+ 5y = 6e
3x
.
SOLUTION As we saw above, if we apply y
2y
+ 5y to z(x) = Ae
3x
we will get
an expression of the form Ke
3x
. We want to determine A so that K = 6. The constant
A is called an undetermined coecient. We have
z = Ae
3x
, z
= 3Ae
3x
, z
= 9Ae
3x
.
94
Substituting z and its derivatives into the left side of the dierential equation, we get
9Ae
3x
2
_
3Ae
3x
_
+ 5
_
Ae
3x
_
= (9A 6A+ 5A)e
3x
= 8Ae
3x
.
We want
z
2z
+ 5z = 6e
3x
,
so we set
8Ae
3x
= 6e
3x
which gives 8A = 6 and A =
3
4
.
Thus, z(x) =
3
4
e
3x
is a particular solution of y
2y
+ 5y = 6e
3x
. (Verify this.)
You can also verify that
y = C
1
e
x
cos 2x + C
2
e
x
sin 2x +
3
4
e
3x
is the general solution of the equation.
If we set z(x) = A cos x and calculate z
and z
, we get
z = A cos x, z
= A sin x, z
=
2
A cos x.
Therefore, L[y] = y
+ay
+az
+ bz =
2
A cos x + a (A sin x) +b(Acos x)
= (
2
A+bA) cos x + (aA) sin x.
That is, L transforms z = Acos x into an expression of the form
Kcos x + M sin x
where K and M are constants which depend on a, b, and A. We will get exactly the
same type of result if we apply L to z = B sinx. Combining these two results, it follows
that L[y] = y
+ay
+by applied to
z = Acos x + Bsin x
will produce the expression
Kcos x + M sin x
where K and M are constants which depend on a, b, , A, and B.
Now suppose we have a nonhomogeneous equation of the form
y
+ay
+ by = c cos x or y
+ ay
+ by = d sin x,
or the general form
y
+ay
+ ay
+ ay
+ 2y
+ y = 10 cos 3x.
SOLUTION Set z = Acos 3x + Bsin 3x where A and B are constants which are to
be determined so that z
+ 2z
+ z = 10 cos 3x.
Calculating the derivatives of z, we have
z = Acos 3x + Bsin 3x, z
+ 2z
4
5
cos 3x +
3
5
sin 3x.
Example 3. Find the general solution of the dierential equation
y
6y
6y
+ 8y = 0.
The characteristic equation is: r
2
6r +8 = (r 2)(r 4) = 0 and the roots are r = 2, 4.
Therefore the functions y
1
(x) = e
2x
, y
2
(x) = e
4x
form a fundamental set of solutions of
the reduced equation.
96
Next we nd a particular solution of (). Since f(x) = 2 cos 2x 4 sin 2x, we seek
a solution of the form z = Acos 2x + B sin 2x. We calculate the derivatives of z and
substitute into the left side of ():
z = Acos 2x + Bsin 2x, z
6z
1
4
cos 2x
1
4
sin 2x.
Continuing with these ideas, if y
+ay
+by is applied to z = Ae
x
cos x+Be
x
sin x,
then the result will have the form
Ke
x
cos x + Me
x
sin x
where K and M are constants which depend on a, b, , , A, B. Therefore, we expect
that a nonhomogeneous equation of the form
y
+ay
+ by = ce
x
cos x +de
x
sin x
will have a particular solution of the form z = Ae
x
cos x + Be
x
sin x.
The following table summarizes our discussion to this point.
A particular solution of y
+ ay
+ by = f(x)
If f(x) = try z(x) =
ce
rx
Ae
rx
c cos x + d sin x z(x) = A cos x +B sin x
ce
x
cos x + de
x
sin x z(x) = Ae
x
cos x + Be
x
sin x
Note: The rst line includes the case r = 0;
if f(x) = ce
0x
= c, then z = Ae
0x
= A.
97
Unfortunately, the situation is not quite as simple as it appears; there is a diculty.
Example 4. Find a particular solution of the nonhomogeneous equation
y
5y
+ 6y = 4e
2x
. ()
SOLUTION According to the table, we should set z(x) = Ae
2x
. Calculating the derivatives
of z, we have
z = Ae
2x
, z
= 2Ae
2x
, z
= 4Ae
2x
.
Substituting z and its derivatives into the left side of (), we get
z
5z
+ 6z = 4Ae
2x
5(2Ae
2x
) + 6(Ae
2x
) = 0Ae
2x
.
Clearly the equation
0Ae
2x
= 4e
2x
which is equivalent to 0A = 4
does not have a solution. Therefore equation () does not have a solution of the form
z = Ae
2x
.
The problem here is z = Ae
2x
is a solution of the reduced equation
y
5y
+ 6y = 0.
(The characteristic equation is r
2
5r+6 = 0; the roots are r = 2, 3; and y
1
= e
2x
, y
2
=
e
3x
are linearly independent solutions.)
In Example 2 of the preceding section we saw that z(x) = 4xe
2x
is a particular
solution of (). So, in the context here, since our trial solution z = Ae
2x
solves the
reduced equation, well try z = Axe
2x
. The derivatives of this z are:
z = Axe
2x
, z
= 2Axe
2x
+Ae
2x
, z
= 4Axe
2x
+ 4Ae
2x
.
Substituting into the left side of (), we get
z
5z
+ 6z = 4Axe
2x
+ 4Ae
2x
5(2Axe
2x
+Ae
2x
) + 6(Axe
2x
)
= Ae
2x
.
Setting z
5z
+ 6z = 4e
2x
gives
Ae
2x
= 4e
2x
which implies A = 4.
Thus, z(x) = 4xe
2x
is a particular solution of () (as we already know).
We learn from this example that we have to make an adjustment if our trial solution z
(from the table) satises the reduced equation. Heres another example.
98
Example 5. Find a particular solution of
y
+ 6y
+ 9y = 5e
3x
. ()
SOLUTION The reduced equation, y
+ 6y
= 3Ax
2
e
3x
+ 2Axe
3x
, z
= 9Ax
2
e
3x
12Axe
3x
+ 2Ae
3x
.
Substituting into the left side of (), we get
z
+ 6z
+ 9z = 9Ax
2
e
3x
12Axe
3x
+ 2Ae
3x
+ 6(3Ax
2
e
3x
+ 2Axe
3x
) + 9(Ax
2
e
3x
)
= 2Ae
3x
.
Setting z
+ 6z
+ 9z = 5e
3x
gives
2Ae
3x
= 5e
3x
which implies A =
5
2
.
Thus, z(x) =
5
2
x
2
e
3x
is a particular solution of ().
The general solution of () is: y = C
1
e
3x
+C
2
xe
3x
+
5
2
x
2
e
3x
.
Based on these examples we amend our table to read:
Table 1
A particular solution of y
+ ay
+ by = f(x)
If f(x) = try z(x) =*
ce
rx
Ae
rx
c cos x +d sin x z(x) = A cos x +B sin x
ce
x
cos x + de
x
sin x z(x) = Ae
x
cos x + Be
x
sin x
*Note: If z satises the reduced equation, try xz; if xz also satises the
reduced equation, then x
2
z will give a particular solution
99
Remark In practice it is a good idea to solve the reduced equation before selecting the
trial solution z of the nonhomogeneous equation. That way you will not waste your time
selecting a z that satises the reduced equation.
Example 6. Without carrying out the calculations, give the form of the general solution
of the nonhomogeneous dierential equation
y
6y = 2 cos x 2e
3x
+ 10.
SOLUTION The rst step is to solve the reduced equation y
6y = 0. The
characteristic equation is r
2
r 6 = (r +2)(r 3) = 0. Thus, y
1
(x) = e
2x
, y
2
(x) = e
3x
forms a fundamental set of solutions of the reduced equation and
y = C
1
e
2x
+ C
2
e
3x
is the general solution of the reduced equation.
Now we need a particular solution z of the given equation. To nd z we make use of
the Corollary to Theorem 3 in the preceding section. That is, well nd a particular solution
z
1
of
y
6y = 2 cos x,
a particular solution z
2
of
y
6y = 2e
3x
,
and a particular solution z
3
of
y
6y = 10.
Then z = z
1
+z
2
+z
3
will be a particular solution of the given equation.
From Table 1, z
1
has the form z
1
(x) = Acos x + Bsin x and z
2
has the form
z
2
(x) = Cxe
3x
. To nd z
3
note that 10 = 10e
0x
which is simply the case r = 0 in the
rst line of Table 1. Thus z
3
has the form z
3
(x) = De
0x
= D. A particular solution z
of the nonhomogeneous equation has the form
z(x) = Acos x +Bsin x + Cxe
3x
+ D.
where the constants A, B, C, D are to be determined.
The general solution of the equation will have the form
y = C
1
e
2x
+C
2
e
3x
+ Acos x + Bsin x + Cxe
3x
+ D.
You can verify that y = C
1
e
2x
+C
2
e
3x
7
25
cos x
1
25
sin x
2
5
xe
3x
5
3
is the general
solution.
100
So far we have only considered the nonhomogeneous dierential equation (1) in cases
where the nonhomogeneous term f is a constant multiple of one of the functions e
rx
, cos x,
sin x, e
x
cos x, e
x
sin x, or is a sum of such functions. In general, the method of
undetermined coecients can be applied in cases where
f(x) = p(x)e
rx
f(x) = p(x) cos x, or p(x) sin x,
f(x) = p(x)e
x
cos x, or p(x)e
x
sin x
where p is a polynomial, or where f is a sum of such functions. This follows from the
fact that the expression y
+ay
+by applied to
z =
_
A
0
+ A
1
x +A
2
x
2
+ + A
n
x
n
_
e
rx
will result in an expression of the form P(x)e
rx
where P is a polynomial of degree n
(or less); y
+ay
+by applied to
z =
_
A
0
+ A
1
x +A
2
x
2
+ + A
n
x
n
_
cos x
will result in an expression of the form P(x) cos x + Q(x) sin x where P and Q are
a polynomials of degree n (or less); and so on.
The general version of the method of undetermined coecients can be summarized as
follows:
(1) If f(x) = p(x)e
rx
where p is a polynomial of degree n, then
z(x) =
_
A
0
+ A
1
x +A
2
x
2
+ + A
n
x
n
_
e
rx
.
(2) If f(x) = p
1
(x) cos x +p
2
(x) sin x where p
1
and p
2
are polynomials of degrees
k and m, respectively, then
z(x) = (A
0
+A
1
x + +A
n
x
n
) cos x + (B
0
+ B
1
x + + B
n
x
n
) sin x
where n = max {k, m}.
(3) If f(x) = p
1
(x)e
x
cos x + p
2
(x)e
x
sin x where p
1
and p
2
are polynomials of
degrees k and m, respectively, then
z(x) = (A
0
+ A
1
x + + A
n
x
n
) e
x
cos x + (B
0
+ B
1
x + + B
n
x
n
) e
x
sin x
where n = max {k, m}.
Note: If any term in z satises the reduced equation y
+ay
+ 4y = (3 + 2x)e
2x
. ()
SOLUTION The functions y
1
(x) = cos 2x, y
2
(x) = sin 2x form a fundamental set of
solutions of the reduced equation y
+ 4y = 0.
A particular solution of () will have the form z = (A + Bx)e
2x
where A and B
are to be determined. The derivatives of z are:
z = (A+ Bx)e
x
, z
= 2(A+Bx)e
2x
+Be
2x
, z
= 4(A+ Bx)e
2x
4Be
2x
.
Substituting z and its derivatives into the left side of (), we get
z
+ 4z = 4(A+ Bx)e
2x
4Be
2x
+ 4(A+Bx)e
2x
= [(8A4B) + 8Bx]e
2x
.
Thus z is a solution of () if
[(8A4B) + 8Bx]e
2x
= (3 + 2x)e
2x
which implies 8A 4B = 3 and 8B = 2.
The solution of this pair of equations is A =
1
2
, B =
1
4
, and
z(x) =
_
1
2
+
1
4
x
_
e
2x
is a particular solution of ().
Example 8. Give the form of the general solution of each of the following nonhomogeneous
equations:
(a) y
3y
+ 2y = (1 + 2x 4x
2
)e
2x
.
(b) y
+ 4y
+ 4y = (3 5x)e
2x
.
SOLUTION (a) The reduced equation is y
3y
+ 4y
2y
3y = 3e
2x
.
2. y
+ 2y
+ 2y = 10e
x
.
3. y
+ 6y
+ 9y = 9e
3x
.
4. y
+ 6y
+ 9y = e
3x
.
5. y
+ 2y
= 4 sin 2x.
6. y
+ 3y
+ y = x
2
+ 3 sin x.
8. y
6y
+ 9y = e
3x
.
9. y
+ 5y
+ 6y = 3x + 4.
10. y
+ 4y
+ 4y = xe
x
.
11. y
+ 6y
+ 8y = 3e
2x
.
12. y
+ 2y
+y = xe
x
.
13. y
+ 9y = x
2
e
3x
+ 6.
14. y
+y
2y = x
3
+ x.
15. y
2y
+ 5y = e
x
sin 2x.
16. y
+ 2y
+ 5y = e
2x
cos x.
Find the solution of the given initial-value problem.
17. y
+y
2y = 2x; y(0) = 0, y
(0) = 1.
18. y
+ 4y = x
2
+ 3e
x
; y(0) = 0, y
(0) = 2.
19. y
(0) = 1.
20. y
2y
+y = xe
x
+ 4; y(0) = 1, y
(0) = 1.
Determine a suitable form for a particular solution z = z(x) of the given equation.
21. y
2y
3y = 6 3xe
x
+ 4 cos 3x.
22. y
+ 2y
= 2x +x
2
e
3x
+ sin 2x.
104
23. y
+y = x
2
1 + 3 cos x 2 sin x.
24. y
5y
+ 6y = 2e
2x
cos x 3xe
3x
+ 5.
25. y
4y
+ 4y = 2xe
2x
+x
2
1 + 2x cos 2x.
26. y
+ 5y
+ 6y = 2e
2x
cos x 3xe
3x
+ 5e
2x
.
27. y
+ 2y
+ 2y = 4e
x
+ 2e
x
cos x + 9.
28. y
+ 2y
+ 5y = 4e
x
sin 2x + 2e
x
cos x.
Find the general solution of the given dierential equation.
29. y
4y
+ 4y = 2 sin x + 3x
1
e
2x
.
30. y
2y
+y =
e
x
x
2
+ 1
+ 2e
2x
.
31. y
+ 9y = 3 cos x 9 sec
2
3x.
32. y
+ 4y = 5e
4x
+ 3 sec
2
2x.
Exercises 33 and 34 are concerned with the dierential equation
y
+ay
+by = f(x)
where a and b are nonnegative constants.
33. Suppose that a, b > 0. Show that if y
1
(x) and y
2
(x) are solutions of the equation,
then y
1
(x) y
2
(x) 0 as x . What happens if a = 0 and b > 0?
34. If f(x) = c, c a constant, show that every solution y(x) of the equation has the
property y(x) c/b as x . What happens if b = 0? What happens if
a = b = 0?
105
3.6 Vibrating Mechanical Systems
Undamped Vibrations
A spring of length l
0
units is suspended from a support. When an object of mass m is
attached to the spring, the spring stretches to a length l
1
units. If the object is then pulled
down (or pushed up) an additional y
0
units at time t = 0 and then released, what is the
resulting motion of the object? That is, what is the position y(t) of the object at time
t > 0? Assume that time is measured in seconds
We begin by analyzing the forces acting on the object at time t > 0. First, there is the
weight of the object (gravity):
F
1
= mg.
This is a downward force. We choose our coordinate system so that the positive direction
is down. Next, there is the restoring force of the spring. By Hookes Law, this force is
proportional to the total displacement l
1
+ y(t) and acts in the direction opposite to the
displacement:
F
2
= k[l
1
+ y(t)] with k > 0.
The constant of proportionality k is called the spring constant. If we assume that the
spring is frictionless and that there is no resistance due to the surrounding medium (for
example, air resistance), then these are the only forces acting on the object. Under these
conditions, the total force is
F = F
1
+ F
2
= mg k[l
1
+ y(t)] = (mg kl
1
) ky(t).
Before the object was displaced, the system was in equilibrium, so the force of gravity,
mg plus the force of the spring, kl
1
, must have been 0:
mg kl
1
= 0.
Therefore, the total force F reduces to
F = ky(t).
By Newtons Second Law of Motion, F = ma (force = mass acceleration), we have
ma = ky(t) and a =
k
m
y(t).
Therefore, at any time t we have
a = y
(t) =
k
m
y(t) or y
(t) +
k
m
y(t) = 0.
When the acceleration is a constant negative multiple of the displacement, the object is
said to be in simple harmonic motion.
106
Since k/m > 0, we can set =
_
k/m and write this equation as
y
(t) +
2
y(t) = 0, (1)
a second order linear homogeneous equation with constant coecients. The characteristic
equation is
r
2
+
2
= 0
and the characteristic roots are i. The general solution of (1) is
y = C
1
cos t +C
2
sin t.
In Exercises 3.6 (Problem 5) you are asked to show that the general solution can be written
as
y = Asin(t +
0
), (2)
where A and
0
are constants with A > 0 and
0
[0, 2). For our purposes here,
this is the preferred form. The motion is periodic with period T given by
T =
2
,
a complete oscillation takes 2/ seconds. The reciprocal of the period gives the number
of oscillations per second. This is called the frequency , denoted by f:
f =
2
.
Since sin(t +
0
) oscillates between 1 and 1,
y(t) = Asin(t +
0
)
oscillates between A and A. The number A is called the amplitude of the motion.
The number
0
is called the phase constant or the phase shift . The gure gives a typical
graph of (2).
t
A
A
y
Figure 1
107
Example 1. Find an equation for the oscillatory motion of an object, given that the period
is 2/3 and at time t = 0, y = 1, y
= 3.
SOLUTION In general the period is 2/, so that here
2
=
2
3
and therefore = 3.
The equation of motion takes the form
y(t) = Asin(3 +
0
).
Dierentiating the equation of motion gives
y
(t) = 3Acos(3t +
0
).
Applying the initial conditions, we have
y(0) = 1 = Asin
0
, y
(0) = 3 = 3Acos
0
and therefore
Asin
0
= 1, Acos
0
= 1.
Adding the squares of these equations, we have
2 = A
2
sin
2
0
+A
2
cos
2
0
= A
2
.
Since A > 0, A =
2.
Finally, to nd
0
, note that
2 sin
0
= 1 and
2 cos
0
= 1.
These equations imply that
0
= /4. Thus, the equation of motion is
y(t) =
2 sin(3t +
1
4
).
Damped Vibrations
If the spring is not frictionless or if there the surrounding medium resists the motion of the
object (for example, air resistance), then the resistance tends to dampen the oscillations.
Experiments show that such a resistant force R is approximately proportional to the
velocity v = y
with c > 0.
Taking this force into account, the force equation reads
F = ky(t) cy
(t).
108
Newtons Second Law F = ma = my
then gives
my
(t) = ky(t) cy
(t)
which can be written as
y
+
c
m
y
+
k
m
y = 0. (c, k, m all constant) (3)
This is the equation of motion in the presence of a damping factor.
The characteristic equation
r
2
+
c
m
r +
k
m
= 0
has roots
r =
c
c
2
4km
2m
.
There are three cases to consider:
c
2
4km < 0, c
2
4km > 0, c
2
4km = 0.
Case 1: c
2
4km < 0. In this case the characteristic equation has complex
roots:
r
1
=
c
2m
+i, r
2
=
c
2m
i where =
4kmc
2
2m
.
The general solution is
y = e
(c/2m)t
(C
1
cos t + C
2
sin t)
which can also be written as
y(t) = Ae
(c/2m)t
sin(t +
0
) (4)
where, as before, A and
0
are constants, A > 0,
0
[0, 2). This is
called the underdamped case. The motion is similar to simple harmonic motion
except that the damping factor e
(c/2m)t
causes y(t) 0 as t .
The oscillations continue indenitely with constant frequency f = /2 but
diminishing amplitude Ae
(c/2m)t
. This motion is illustrated in Figure 2.
t
y
Figure 2
109
Case 2: c
2
4km > 0. In this case the characteristic equation has two distinct
real roots:
r
1
=
c +
c
2
4km
2m
, r
2
=
c
c
2
4km
2m
.
The general solution is
y(t) = y = C
1
e
r
1
t
+C
2
e
r
2
t
. (5)
This is called the overdamped case. The motion is nonoscillatory. Since
_
c
2
4km <
c
2
= c,
r
1
and r
2
are both negative and y(t) 0 as t .
Case 3: c
2
4km = 0. In this case the characteristic equation has only one
real root:
r
1
=
c
2m
,
and the general solution is
y(t) = y = C
1
e
(c/2m) t
+C
2
t e
(c/2m) t
. (6)
This is called the critically damped case. Once again, the motion is nonoscillatory
and y(t) 0 as t .
In both the overdamped and critically damped cases, the object moves back to the
equilibrium position (y(t) 0 as t ). The object may move through the equilibrium
position once, but only once. Two typical examples of the motion are shown in Figure 3.
t
y
t
y
Figure 3
Forced Vibrations
The vibrations that we have considered thus far result from the interplay of three forces:
gravity, the restoring force of the spring, and the retarding force of friction or the surround-
ing medium. Such vibrations are called free vibrations .
110
The application of an external force to a freely vibrating system modies the vibrations
and produces what are called forced vibrations . As an example well investigate the eect
of a periodic external force F
0
cos t where F
0
and are positive constants.
In an undamped system the force equation is
F = kx + F
0
cos t
and the equation of motion takes the form
y
+
k
m
y =
F
0
m
cos t.
We set =
_
k/m and write the equation of motion as
y
+
2
y =
F
0
m
cos t. (7)
As well see, the nature of the motion depends on the relation between the applied frequency
, /2, and the natural frequency of the system, /2.
Case 1: = . In this case the method of undetermined coecients gives the
particular solution
z(t) =
F
0
/m
2
cos t
and the general equation of motion is
y = Asin(t +
0
) +
F
0
/m
2
cos t. (8)
If / is rational, the vibrations are periodic. If / is not rational, then
the vibrations are not periodic and can be highly irregular. In either case, the
vibrations are bounded by
|A| +
F
0
/m
2
2
.
Case 2: = . In this case the method of undetermined coecients gives
z(t) =
F
0
2m
t sin t
and the general solution has the form
y = Asin(t +
0
) +
F
0
2m
t sin t. (9)
The system is said to be in resonance . The motion is oscillatory but, because of
the t factor in the second term, it is not periodic. As t , the amplitude
of the vibrations increases without bound.
111
A typical illustration of the motion is given in Figure 4.
t
y
Figure 4
Exercises 3.6
1. An object is in simple harmonic motion. Find an equation for the motion given that
the period is
1
4
and, at time t = 0, y = 1, y
2
cos t.
112
10. Show that if / is rational, then the vibrations given by (8) are periodic.
11. Show that if = , then the method of undetermined coecients applied to (7)
gives
z =
F
0
2m
t sin t.
113